J . Med. Chem. 1988, 31, 384-389
384
Toward the Synthesis of Isozyme-Specific Enzyme Inhibitors. Potent Inhibitors of Rat Methionine Adenosyltransferases. Effect of One-Atom Elongation of the Ribose-Pa Bridge in Two Covalent Adducts of L-Methionine and P,y-Imido-ATP’ Francis Kappler, Vivekananda M. Vrudhula, and Alexander Hampton” Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsyluania 19111. Received April 10, 1987 With 2’,3’-0-isopropylideneadenosine or its NG-benzoyl derivative as starting material, synthetic routes to two novel adducts of L-methionine and P,y-imido-ATP have been devised. One adduct, 14 (2:3 mixture of 6’ epimers), had a P”OCH(R’)CH, system [R’ = CH2-~-SCH2CH,CH(NH2)C02H] in place of the PaOC(5’)Hz system of ATP, while the other, 16 (2:3 mixture of 5’ epimers), had a PaOCH2CH2CH(Rz)system [R2= L-SCH,CH,CH(NH,)CO~H]. The ribose-P“ bridge in 14 and 16 contained one more methylene group than in two homologous methionine-ATP adducts studied previously. Adduct 14 was a potent inhibitor of the rat M-2 (normal tissue) and M-T (Novikoff ascitic hepatoma) variants of methionine adenosyltransferase and gave competitive kinetics vs MgATP (Ki = 0.39 and 0.63 pM, respectively) or vs L-methionine (Ki = 2.2 and 2.7 pM). Adduct 16 was likewise a potent inhibitor competitive vs MgATP (Ki = 0.44 and 0.81 pM, respectively) or L-methionine (Ki = 2.1 and 1.5 WM).The kinetic data indicate that 14 and 16 inhibit by binding simultaneously to the MgATP and L-methionine substrate sites and that the extra methylene group facilitates the interaction of their methionine residues with these methionine sites.
On grounds outlined previously,1,2enzyme inhibitors selective for certain isozymic variants predominant in isozyme profiles of tumor tissues have the potential to be more toxic toward poorly differentiated malignant tissue than toward many normal tissues of the host species and are thus of interest as possible progenitors of new antineoplastic drugs. The present series of studies is examining approaches that might be useful in the systematic design of potent, isozyme-selective inhibitors. Selected as model targets are mammalian isozymes that catalyze reactions involving two substrates. The work has revealed a three-stage strategy that can generate potent and isozyme-selective inhibitors relatively r e a d i l ~ . ~The , ~ scope and limitations of this strategy are as yet unclear. The second stage comprises attempted derivation of a potent dual substrate-site (“multisubstrate”) inhibitor by the known procedure5 of covalently coupling two substrates via atoms involved in the enzyme-catalyzed reaction. In the third stage, one or more substrate substituents that, from studies in the first stage, are known to produce isozyme-selective inhibitors are attached in the corresponding positions to adducts generated in the second ~ t a g e . ~ , ~ The present paper describes further studies of the second of the above stages. Earlier studies found that twosubstrate adducts of the above type could act as potent inhibitors of rat isozymes of thymidine kinase6 and adenylate k i n a ~ ethough ,~ not of hexokinase.6 More recent papers reported that the 5’R epimers of adducts of Lmethionine with @,y-imido-ATP1or with a 5’-(aminomethyl)-5’-deoxy-ATP analogue’ (13 and 15, respectively) This is part 15 in a series on the design of isozyme-specific enzyme inhibitors. For part 14, see: Kappler, F.; Vrudhula, V. M.; Hampton, A. J . Med. Chem. 1987, 30, 1599. Hampton, A.; Kappler, F.; Maeda, M.; Patel, A. D. J . Med. Chem. 1978,21, 1137. Hampton, A.; Kappler, F.; Picker, D. J . Med. Chem. 1982,25, 638.
Kaualer. F.: Hai. T. T.: Abo. M.: Hamuton, A. J . Med. Chem. 1982,25, 1179. (a) Lienhard. G. E. Annu. Reo. Med. Chem. 1972. 7. 249. (b) Wolfenden, R. Acc. Chem. Res. 1972,5, 10. (c)Lienhard, G. E.; Secemski, I. I. J . Biol. Chem. 1973,248, 1121. (d) Tomasselli, A. G.; Schirmer, R. H.; Noda, L. H. Eur. J. Biochem. 1979,93, 257. Hampton, A.; Hai, T. T.; Kappler, F.; Chawla, R. R. J . Med. Chem. 1982,25, 801. Vrudhula, V. M.; Kappler, F., Hampton, A. J . Med. Chem. 1987, 30, 888.
are potent inhibitors of the M-T variant (predominant in rat ascitic hepatoma cells) and the M-2 variant (predominant in most rat normal tissues) of methionine adenosyltransferase. This enzyme catalyzes attack of methio0
0 0
HO OH
13 “2
7’
I
HO
I
OH
14
nine sulfur on C5’ of ATP with expulsion of tripolyphosphate. As discussed earlier,s the M-T variant is of interest as a model mammalian target in the attempted design of an isozyme-selective type of antineoplastic agent. The present study explores the effect on inhibitory potency of ribose-P* bridge elongation brought about by insertion of an additional CH, between C4’ and C5’ of 13, giving 14, and between C5’ and C6’ of 15, giving 16. Adduct 16 also differs from 15 in possessing a POCHz system, which is hydrolytically more stable at physiological pH than the PNHCH, system of 15 and in possessing a p,y-imido group which, in P,y-imido-ATP, imparts phosphatase inertnessg and hence may render 16 more stable against such attack in vivo. Moreover, in structurally related methionine-ATP adducts a p,y-imido group was found to increase inhibitory potency ca. 2.5-f01d.~’ Adduct 14 was synthesized and evaluated here as a mixture, in known proportion, of its 6’ epimers, and adduct 16 was synthesized and evaluated (8) Kappler, F.; Hai, T. T.; Hampton, A. J . Med. Chem. 1986,29, 318. (9) Yount, R. G. Aduan. Enzyrnol. Relat. Areas Mol. Biol. 1975, 43, 1. (10) Kappler, F.; Hai, T. T.; Cotter, R. J.; Hyver, K. J.; Hampton, A. J . Med. Chem. 1986,29, 1030.
0022-262318811831-0384$01.50/0@ 1988 American Chemical Society
Two Adducts of L-Methionine and @,y-Imido- ATP
Journal of Medicinal Chemistry, 1988, Vol. 31, No. 2 385 Scheme I
as a mixture of its 5’ epimers. “2
0 0
X
2: R = H 3: R=CHzOH
1 Ad=adenin- 9 - y l HO OH
HO-
15 “2
0
0 0
I1 I I I I (H0)zPNHPI OI
-
2. 1. TiCIj, MeONap H 5
P O C b C G
J.!,
N/I
0%“”””’ 0 0
x
4 HO OH HQC( NH 2)CHCH2CH2S HO OH
16
Synthesis of 14. The readily accessible P-benzoyl5’-deoxy-5’-(nitromethyl)-2’,3‘-O-isopropylideneadenosine (1;Scheme 1),l1when treated in t-BuOH with 1.2 equiv 0 0 of CHzO in the presence of 1 equiv of t-BuOK, gave a x mixture of products from which the desired 6’-hydroxymethyl derivative 2 could be isolated in homogeneous form in 48% yield. HPLC and lH NMR analyses indicated that the two 6’ epimers of 2 were present in a 1:l ratio. Also isolated from the reaction mixture were a more polar UV-absorbing product (30%), presumably the 6’,6’-bis(hydroxymethyl) derivative 3, and unreacted 1 (10%). Platinum-catalyzed hydrogenation of 2 gave only low yields of a product that appeared to be the amine 7. In a successful alternate approach to the desired diol 5, 2 was 0 0 converted in MeOH to a sodium nitronate, and this was reduced with T i c & a t pH 5 by the McMurry-Melton x procedure12 to give, via hydrolysis of the corresponding 6 imine, the 6’-keto nucleoside 4 in about 60% yield. Also formed were NG-benzoyladenine (10% yield) and 1 (10% ) . The a-amino acid functionality in 9 was protected by Treatment of crude 4 with NaBH, in MeOH brought about conversion to an N-Boc derivative of a methyl ester (10) reduction of the keto group as well as debenzoylation a t by means of successive treatments with di-tert-butyl pyN6 and furnished homogeneous 5 in 27 % yield overall from rocarbonate and CH2N2. This furnished homogeneous 10 2. A major byprodud was adenine. Elimination of adenine in 62% yield from 9 with an unaltered ratio of 6’ epimers. in this reaction and of M-benzoyladenine in the reduction By use of procedures described previously in the synthesis of 2 presumably proceeds through a 6’ carbanion and is of 13,l 10 was converted into the nucleoside 6’-phosphate accompanied by formation of a 4’3’ double bond. The derivative 11 (64% yield), and this in turn was converted same mechanism has been proposed for base-catalyzed to its imidopyrophosphoroxy derivative 12 (58% yield), elimination of adenine from 5’-cyano-5’-deoxyadenosine.13 after which removal of blocking groups afforded the hoIn confirmation of their structure, a mixture of the two mogenous adduct 14 (29% yield from 10) as a 2:3 mixture epimeric diols 5 , upon treatment with NaIO,, yielded a of the two 6’ stereoisomers as indicated by the observation, single aldehyde, which, after reduction with NaBH,, was from HPLC analysis of unreacted 10,that the two 6’ steconverted to an alcohol that could not be distinguished reoisomers of 10 had undergone phosphorylation a t the from authentic 9-(5-deoxy-2,3-O-isopropylidene-~-~-allo-same rate. Adduct 14 was isolated as a tetrasodium salt furanosyl)adenine14(6)by reversed-phase HPLC. The diol that possessed the expected elemental analyses and UV, 5 was converted quantitatively to a mono-0-tosylate that, lH NMR, and 31PNMR spectral properties. presumably, was predominantly or exclusively the primary Synthesis of 16. P-Benzoyl-2’,3’-0-isopropylidenetosylate 8. Without purification, this was treated in EtOH adenosine-5’-aldehyde, conveniently obtained through at 22 “C with disodium L-homocysteinate. The 6‘-hydroxy oxidation of the corresponding alcohol,15 was condensed, 7’-thioether nucleoside derivative 9 formed as a 1:l mixture without isolation, with (carbethoxymethy1ene)triphenylof the two 6’ epimers and after reversed-phase chromaphosphorane to give, after preparative HPLC, the trans tography was isolated in homogeneous form in 77% yield deisomer of the 5’-deoxy-5’-~carbethoxyvinyl)adenosine overall from 5 as a 2:3 mixture of 6‘ epimers. rivative 17a (Scheme 11) in 73% yield. Treatment of 17a in aqueous THF at pH 9 with a two-fold excess of sodium L-homocysteinate gave a 4:5 mixture of the 5’ epimers of (11) Mock, G. A,; Moffatt, J. G. Nucleic Acids Res. 1982,10, 6223. the 5’-(alkylthio)-5’-deoxyadenosinederivative Ma,isolated (12) McMurry, J. E.; Melton, J. J. Org. Chem. 1973, 38, 4367. (13) Meyer, W.; Follmann, H. Chem. Ber. 1980, 113, 2530.
HowAd
(14) Kappler, F.; Hampton, A. Synthetic Procedures in Nucleic Acid Chemistry, Vol. 4; Townsend, L. B., Tipson, R. S., Eds.; Wiley: New York; in press.
(15) Ranganathan, R. S.; Jones, G. H.; Moffatt, J. G. J. Org. Chem. 1974, 39, 290.
386 Journal of Medicinal Chemistry, 1988, Vol. 31, No. 2
Scheme I1
Table I. Inhibition Constants of Adenine Nucleotide Derivatives with Kidney (M-2) and Novikoff Ascitic Hepatoma (M-T) Forms of Rat Methionine Adenosvltransferase" M-T: Ki, KM (type of M-2: K), 1 M (type of inhibn)* inhibn) com.od ATP varied Met varied ATP varied Met varied 13c 0.13 (C) 0.65 (C) 0.21 (C) 0.67d (M) 14 0.39 (C) 2.2 (C) 0.63 (C) 2.7 ( C ) 15' 0.36 (C) 3.0 (NC) 0.27 (C) 2.3 (NC) 16 0.44 (C) 2.1 (C) 0.81 (C) 1.5 (C) When methionine (Met) was the variable substrate, [MgATP] was 2 mM with both MAT forms; with variable MgATP, [Met] was 60 r M with M-2 and 120 1M with M-T. For other conditions, see the Experimental Section. The Michaelis constant (KM) of MgATP was 0.14 mM with both M-2 and M-T. The KM of Lmethionine was 4.4 pM with M-2 and 12 1M with M-T under the conditions of the K, determinations. * C = competitive; NC = simple noncompetitive (the inhibitor reduces V, and does not change KM);M = mixed C and NC. CDatafrom ref 1. dCalculated from double-reciprocal plots (see the Experimental Section) as K , (slope); K , calculated from intercepts on the 1/V axis was 1.75 pM. 'Data from ref 7.
NHR
I
17a: R=COPh 17b: R z H
0 0
x
18a: R = COPh 18b: R = H "2
I
..-i-;-l'J I
Kappler et al.
,
0 0
X
19: R'=L-SCH2CH2CH(NH2)COzHI R2=OH 20: R'= L-SCH2CH2CH(NH-Boc)C02MeI R2=OH 21: R'= L - S C H ~ C H ~ C H ( N H - B O C ) C O ~R2=OP(0)(OH)2 M~, 22: R1= L - S C H ~ C H ~ C H ( N H - B O C ) C O ~ M ~ , R2; OP(OH)(O)OP(OH)(O)NHP(O)(OH):!
in 82% yield after reversed-phase chromatography. Selective reduction of the carbethoxy group to the corresponding alcohol was attempted with LiBH, in THF16 or with NaBH, in MeOHl' or EtOH.6 These reactions, however, produced a variety of products that appeared to arise, at least partly, from loss of the benzoyl group at N6 and/or elimination of the homocysteine residue. Reaction of the known olefin 17b1*with sodium L-homocysteinate gave the homogeneous mixed 5' epimers of the 5'-(alkylthio)-5'-deoxy nucleoside 18b in 80% yield and 4.5:5 ratio. Treatment of 18b with NaBH, in EtOH yielded mixtures of the desired primary alcohol 19 with unchanged 18b,free homocysteine, and unidentified UV-absorbing material that was less polar than 18b or 19. A study of the effect of temperature and the number of equivalents of NaBH, revealed that a temperature of 60 "C and 4-6 equiv of the NaBH, were roughly optimal for 19 formation and furnished homogeneous 19 in 38% yield. The mixed 5' epimers (in 2:3 ratio) of the methyl ester 20 of the N-Boc derivative of 19 were then prepared in 80% yield from 19 by the methods used to obtain 10. By the procedures used in the above synthesis of 14,20was converted in 60% yield into the corresponding nucleoside 7'-monophosphate derivative 21,and 21,via its imidopyrophosphoroxy derivative 22,was converted into 16 (37% yield of homogeneous material from 21). Adduct 16 was isolated as a tetrasodium (16) Brown, H. C.; Narasimhan, S.; Choi, Y. M. J. Org. Chem. 1982, 47, 4702. (17) Brown, M. S.; Rappoport, H. J . Org. Chem. 1963, 28, 3261. (18) Kim, K. S.; Kang, C. M.; Kim, S. J.; Jung, K. S.; Hahn, C. S. Bull. Korean Chem. SOC.1985, 6 , 350.
salt with the expected UV and 31PNMR spectral properties. Anion-exchange HPLC analysis on a reversed-phase column showed the ratio of 5' epimers in 16 to be the same (2:3) as in its blocked nucleoside precursor 20. Enzyme Studies. Listed in Table I are inhibition constants of 13-16 with M-2 and M-T determined under the same conditions for each of these methionine adenosyltransferase variants. Adduct 14 was a potent inhibitor of both variants and produced competitive kinetics vs MgATP [K,(ATP)/Ki = 360 and 220 for M-2 and M-T, respectively] and vs L-methionine [KM(Met)/Ki= 2.0 and 4.5, respectively], suggesting that it may bind simultaneously to the MgATP and methionine enzymic sites in exerting these inhibitions. Inasmuch as one 5' epimer of 15 inhibits M-T 74-fold more effectively than the other 5' epimer,' it can be expected that the observed inhibitions by 14 are brought about mainly by one of its two 6' epimers (present in a 2:3 ratio). Table I shows that if inhibition arose from the minor epimeric component, the potency of 14 for both enzyme forms would be almost equal to that of 13;if arising from the major epimer, the potency would be 33% less. Adduct 14 inhibited M-T competitively with respect to methionine, whereas 13 gave a mixed competitive-noncompetitive type of inhibition, suggesting that the additional methylene group may permit the methionine residue in 14 to interact with the methionine binding site in a way that diminishes the ability of methionine to bind reversibly to an enzyme-adduct complex to form a ternary methionine-enzyme-adduct complex. Adduct 16,tested as the 2:3 mixture of its 5' epimers, is seen from Table I to inhibit M-2 significantly more effectively than the structurally related adduct 15 irrespective of which 5' epimer of 16 is the more inhibitory. Adduct 16 also inhibited M-T more effectively than did 15 as indicated by its inhibition constant with variable methionine and appeared from its inhibition constant with variable MgATP to be either as effective or almost as effective against M-T as 15,depending upon whether the minor or the major 5' epimer of 16 is the more active species. That adduct 16 inhibited both variants competitively with respect to either methionine or to MgATP suggests that it could be acting as a dual substrate site inhibitor. A P,y-imidophosphate system in a homocysteine-ATP adduct was earlier reported to enhance inhibitory potency toward M-2 and M-T by a factor of ca. 2.5-fold," and hence this system might enhance the potency of 16 also. Notably, with both enzyme variants, 16
Two Adducts of L-Methionine and p,y-Imido- ATP inhibits in competitive fashion with respect to methionine, whereas 15 inhibits noncompetitively (Table I). Possibly, the longer PW5‘ bridge of 16 allows its methionine residue to interact with a larger segment of the methionine binding site, thereby hindering binding of methionine t o the enzyme-16 complex and suppressing noncompetitive inhibition. The ATP moiety of enzyme-bound 16 could, in the manner of ATP itself,lg be bound mainly via its tripolyphosphate and 2’-OH loci as a result of rotations that can occur about the C2’-02’ bond and about bonds i n t h e P”-C4’ system. In summary, previous studiesl~’have described two types of m e t h i o n i n e A T P adducts (13 and 15) that were potent inhibitors of the M-2 and M-T variants and that appeared t o possess a dual substrate site mode of binding. The present study provides evidence that introduction of a single methylene group into the Pa-C5’ bridge of 13 or 15 (concomitantly with two isosteric atom replacements in the case of 15) is compatible with maintenance of powerful dual-site inhibition and appears to permit the methionine moieties t o interact more effectively with the enzymic methionine sites, The four potent inhibitors 13-16 exhibit no more than minor levels of isozyme selectivity. It is possible, however, that selectivity for t h e chemotherapeutic target M-T may be inducible by structural modification, for example by the introduction of substituents indicated to be promising from previous studies with derivatives of ATP and of methionine.8,20 Studies of additional potent and/or selective M-T inhibitors are continuing.
Experimental Studies Chemical Synthesis. General Procedures, P-Toluenesulfonyl chloride was recrystallized from hexane. S-Benzyl-Lhomocysteine was prepared and purified as describedz1 and converted to disodium L-homocysteinate as described.2z H20 used in reactions was degassed. Thin-layer chromatography (TLC) was run on Merck silica gel 60 F-254 plates in (A) CH3COCH3-CHZCl2(37), (B) MeOH-CHC13 (1:19),(C) MeOH-CHC13(1:9) (D) CHC13-MeOH-4% aqueous CH3COOH (3:2:1, lower layer), (E) MeOH-CHC1, (3:7), and (EE) acetone-CH2C12(1:4). Reversed-phase TLC was carried out on glass plates of silica gel RP-18 from EM Laboratories in (Rl) MeOH-H20 (1:l)and (R2) MeOH-H20 (7:3). Paper chromatography was carried out by the ascending technique on Whatman No. 1 in (F) l-propanolNH40H-H20 (55:10:35) and (G) isobutyric acid-1 M NH40H (6:4). Electrophoresis was carried out on Whatman No. 1paper at pH 3.5 (0.05 M citrate) and pH 7.5 (0.05 M Et3NH.HC03). HPLC was performed on a Waters Model 204 chromatograph equipped with a Waters 740 data module, a dual solvent-delivery system (Model M-6000A), a Model 660 programmer, and a Waters RCM-100 radial compression unit containing a 4-pm particle size Nova-Pak CI8 cartridge (8 mm X 10 cm). Gradients were linear and were run for 10 rnin at 2 mL/min flow rate in (H) 40-100% MeOH in HzO, (J) 20-100% MeOH in HzO, and (K) 0-30% MeOH in 0.1 M potassium phosphate-0.025 M (n-Bu),NHS04 (pH 7.6). Flash chromatographyz3was performed with Merck silica gel 60 (230-400 mesh). Baker bonded-phase octadecyl silica gel (flash chromatography grade) was used for column chromatography. Preparative HPLC was performed at 250 mL/min on a Waters Prep LC/System 500A equipped with one silica gel cartridge. UV spectra were obtained with a Varian Model 635 spectrophotometer, and ‘H and 31PNMR spectra were obtained with a Nicolet NT-300 WB spectrometer. Chemical shifts are given as ppm (6) downfield from SiMe4or 85% H3P04. The sign of 31Pshifts are in accord with the 1976 IUPAC c~nvention.’~ (19) Kappler, F.; Hai, T. T.; Hampton, A. Bioorg. Chem. 1985,13, 289. (20) Lim, H.; Kappler, F.; Hai, T. T.; Hampton, A. J. Med. Chem. 1986,29, 1743. (21) Armstrong, M. D. Biochem. Prep. 1957,5, 92. (22) Riegel, B.; du Vigneaud, V. J . Biol. Chem. 1935, 112, 149. (23) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978,43,2923.
Journal of Medicinal Chemistry, 1988, Vol. 31, No. 2 387 Elemental analyses were by Galbraith Laboratories, Inc., Knoxville, TN, and were within *0.4% of theoretical values unless stated otherwise. Compounds were dried in vacuo at 22 “C over P20, except for 2, 5, and 9 which were dried at 78 “C. N6-Benzoyl-5’-deoxy-5’-(2-hydroxy-l ( R S )-nitroethyl)2’,3’-0-isopropylideneadenosine(2). To a stirred solution of 1 (2.33 g, 5.1 mmol) in t-BuOH (25 mL) under argon was added 0.2 M t-BuOK in t-BuOH (25.5 mL, 1 equiv). The suspension was stirred for 15 min. Gaseous HCHO generated from paraformaldehyde (185 mg, 1.2 equiv) was added, and the suspension was stirred for an additional 0.5 h. Solid NH4Cl (1g, 18.7 mmol) was added, t-BuOH was removed in vacuo, and the residue was partitioned between EtOAc (2 X 50 mL) and saturated aqueous NH,Cl. The EtOAc was dried (Na2S04)and evaporated in vacuo. The residue was subjected to flash chromatography on silica gel with acetone-CH2C12 (3:7) to give 2 (1:lmixture of 6’ epimers by NMR H-1’ integration; 1.21 g, 48% from I), presumed 3 (0.78 g, 30%), and 1 (0.23 g) as white foams homogeneous by silica gel TLC [Rfvalues, solvent A, 0.55 (I), 0.30 (Z),0.13 (3); solvent B, 0.42 (I),0.26 (2), 0.19 (3)] and HPLC [solvent H; tR 7.8 min, 8.1 min (2); 8.5 rnin (1); 7.8 rnin (3)]. 2: UV max (EtOH) 280 nm (e 18600); ‘H NMR (CDC13)6 9.10 (br s, 1,NHCO), 8.78, 8.75 ( 2 s, 1, I4-2 or H-8), 8.01 (m, 3, H-2 or H-8 and 2 H of C6H5),7.53 (m, 3, C6H6),6.07, 6.06 (2 d, 1,H-1’, Jl,2, = 1.9, 1.2 Hz), 5.52, 5.50 = 6.3, 6.3 Hz, J3,,,,= 3.6, (m, H-2’), 5.00, 4.97 (2 dd, 1, H-3’, J21,31 3.5 Hz), 4.61,4.58 (m, 1,H-4’), 4.29 (m, 1,H-69, 3.85 (m, 2, H-79, 2.41 (m, 2, H-59, 1.58, 1.37, 1.36 (3 s, 6, C(CH&). Anal. (C22H2*N607.0.75H20) C, H, N. 5’-Deoxy-5’-( 1 ( R S ),2-dihydroxyethy1)-2’,3’-0 -isopropylideneadenosine (5). To a stirred solution of 2 (0.97 g, 2 mmol) in MeOH (16.4 mL) at 20 “C under argon was added 1.5 M NaOMe in MeOH (2.7 mL, 2 equiv). After 2 min, a solution (16.4 mL) prepared from TiC13 (1.54 g), H 2 0 (7.7 mL), and 4 M aqueous NH40Ac (15 mL) under argon was added, and the mixture was stirred for 20 min. MeOH was removed in vacuo, and to the white suspension was added saturated aqueous NH4Cl (30 mL) and EtOAc (50 mL). The mixture was vigorously stirred for about 3 min and then filtered through a bed of Celite (1 X 5 cm) to remove titanium salts. The Celite was washed with EtOAc (3 X 50 mL). The aqueous phase in the combined filtrate and washings was extracted with EtOAc (50 mL). The combined EtOAc solutions were dried (Na2S04)and evaporated in vacuo. Toluene (2 X 25 mL) was evaporated from the residual crude 4, and this was then dissolved in MeOH (10 mL) and treated with NaBH4 (115 mg, 3 mmol). When effervescence ceased, the solution was refluxed for 0.5 h. MeOH was evaporated in vacuo, and the residue was partitioned between CHzClz(30 mL) and water (3 x 10 mL). The combined aqueous solutions were concentrated to 5 mL and applied to a column (3 X 3 cm) of (&bonded silica gel, and this was eluted with water (100 mL) and then with MeOH-HzO (3:7,200 mL). The MeOH-H20 eluate was evaporated in vacuo, and the pale yellow gum was triturated with CH2C12 (2 x 5 mL). The gum obtained after decantation of the CH2C12 was dried in vacuo to give 5 (8130 ODzsounits, 0.54 mmol, 27% yield from 2) as a white foam that was homogeneous by TLC (solvent C, Ri of the two 6’ epimers 0.15,0.13) and HPLC (solvent J, tR 7.0 min): lH NMR (CD3CN) 6 8.19, 7.99, 7.97 (3 s, 2, H-2, H-8), 6.03, 6.02 (2 d, 1, H-1’, J1,2t = 3.2 and 2.7 Hz), 5.39 (m, 1, H-2’), 4.94, 4.86 ( 2 dd, 1,H-3’, J2,,3, = 7.5, 6.3 Hz and J3t,4, = 3.5 and 3.4 Hz), 4.32 (m, 1,H-4’), 3.60 (m, 1,H-6’), 3.29 (m, 2, H-79, 1.69 (m, 2, H-5’), 1.53, 1.31 (2 s, 6, C(CH3),). Integration of H-1’ revealed the 6’ epimers to be in a 1:l ratio. Anal. (C15H21N50&H30H*0.5H,O) C, H, N. 5’-Deoxy-5’-(2-~-homocystein-S -yl-1 ( R S )-hydroxyethyl)-2’,3’-0 -isopropylideneadenosine(9). A solution of 5 (7800 OD260 units, 520 pmol) and p-CH3C6H4S02Cl(158 mg, 1.6 equiv) in anhydrous pyridine (0.9 mL) was kept at -15 “C for 18 h, at which time silica gel TLC revealed quantitative conversion to 8 (Rfin solvent C, 0.50). The pyridine was evaporated in vacuo. A solution of the residue in EtOAc (20 mL) was washed with saturated aqueous NaHC03 (2 X 5 mL), dried (Na2S0,), and evaporated in vacuo. EtOH (10 mL) was evaporated in vacuo from the residue. A solution of the residue in EtOH (1 mL) was (24) Pure Appl. Chem. 1976, 45, 211.
388 Journal of Medicinal Chemistry, 1988, Vol. 31, No. 2 treated with 0.52 M disodium L-homocysteinateZ2in EtOH (1.5 mL, 1.5 equiv), and the orange solution was kept at ambient temperature for 2 h. EtOH was evaporated in vacuo, water (3 mL) was added to the residue, and the resulting turbid solution was adjusted to pH 6 and applied to a C18 silica gel column (3 X 3 cm), which was eluted in succession with HzO (100 mL), MeOH-H20 (L9, 50 mL), MeOH-H,O (3:7, 100 mL), and MeOH-H,O (Ll, 50 mL). The aqueous 30% MeOH eluate was evaporated to dryness in vacuo, and the residue, after trituration with acetone (15 mL), was dried to give 9 (6000 OD260units, 0.4 mmol, 77% from 5) as a white foam homogeneous by TLC (R, 0.10 in D; 8 had Rf 0.80) and HPLC ( t 8.3, ~ 8.6 rnin in J; 8 had t~ 9.8, 10.0 rnin): ‘H NMR (DzO)6 8.12, 8.10, 8.06, 8.05 (4 s, 2, H-2, H-8), 6.08, 6.04 (2 d, 1, H-l’, J1,,2t = 1.2, 2.5 Hz), 5.63, 5.43 (2 dd, 1,H-2’, J2,,3,= 6.1,6.3 Hz), 4.91 (m, 1H-3’), 4.33 (m, 1,H-4’), = 3.60, 3.25 (m, 2, H-6’ and H-2”), 2.46, 2.42 (2 t, 2, H-4”, J3,,,4,8 7.3, 7.6 Hz), 2.20 (m, 2, H-7’), 1.75 (m, 4, H-5’, H-3”), 1.46, 1.28, 1.26 (3 s, 6, C(CH,),). Integration of H-1’ revealed the 6’ epimers C, H, to be in a 2:3 ratio. Anal. (Cl9HzsN6O6S.0.75CH3oH~Hz0) S; N: calcd, 16.47; found, 15.90. The N-tert -Butyloxycarbonyl Derivative of 9, Methyl Ester 10. Compound 9 (94 mg, 0.20 mmol) was dissolved in dry DMF (2 mL) containing E5N (32 pL, 230 pmol), and di-tert-butyl pyrocarbonate (52 mg, 238 ymol) was added.25 After 0.5 h at 22 “C, TLC analysis (E) showed that conversion to a single ninhydrin-negative spot, R f 0.40, was complete. The DMF was evaporated. The residue was dissolved in MeOH (5 mL) and treated with excess of ethereal CH,N2. Volatiles were removed in vacuo, and the residue was eluted from a column (2 X 25 cm) of silica gel with CHC1,-MeOH (97:3). Evaporation of eluant gave 10 as a white foam (86 mg, 74%) homogeneous by TLC, R f 0.40 in C, and by HPLC, tR 9.0 and 9.3 min in H (2:3 ratio of 6’ stereoisomers): NMR (CDCl,) 6 1.35, 1.38, 1.59 (6 H, CMe,), 1.40, 1.43 (9 H, CMe,), 1.95 (4 H, H-5’, H-2”), 2.40 (2 H, H-49, 2.58, (2 H, H-3”), 3.58 (1H, H-6’), 3.72 (3 H, OMe), 3.96 (1 H, H-1”), 4.52 (1H, H-4’), 4.93, 5.18 (1H, H-3’), 5.40, 5.45 (1H, H-2’), 5.80, 5.89 (1 H, H-l’), 7.91, 7.94, and 8.32 (2 H, H-2, H-8). Anal. (C26H3sN608SCH3OH) C, H, N, S. 5’-Deoxy-5’-[[ 1 ( R S )-[ (P,y-imido)triphosphoroxy]-2-~homocystein-S-yl]ethyl]adenosine(Mixture of Diastereomers) (14). Compound 10 (120 pmol) was converted to the homogeneous, purified triethylammonium salt of 14 in 29% yield overall by the procedures used in the synthesis of 13.’ To a solution of this in MeOH (1mL) was added methanolic NaI (180 pL of 1M) followed by acetone (75 mL). The resulting precipitate was washed with acetone and dried in vacuo to give 26 mg of tetrasodium 14 as a white powder, which was homogeneous by paper chromatography in systems (F) and (G), Rf 0.34 and 0.40, respectively: UV max pH 2, 260 nm ( e 14700), pH 12, 260 nm (e 14800); ,‘P NMR (121.46 MHz, DzO) 6 -11.58 (P“),-6.50 (Pb), -1.47 (Py). The lH NMR spectrum in D,O was consistent with the structure, but signal assignments could not be made due to complex overlapping resonances in the two 6’ stereoisomers. Anal. (C16H24N7014P3S Na4.3.5Hz0) C, H, S, P; N calcd, 11.98; found, 12.45. N6-Benzoyl-5’-deoxy-5’-( trans -carbethoxyvinyl)-2’,3’-0 isopropylideneadenosine (17a). To a solution of IP-benzoyl2’,3’-0-isopropylideneadenosine(4.1 g, 10 mmol) in Me,SO (50 mL) were added N,”-dicyclohexylcarbodiimide (8.2 g, 40 mmol) and dichloroacetic acid (0.7 mL, 8.4 mmol). The mixture was stirred for 18 h. The precipitate of N,N’-dicyclohexylurea was removed by filtration and washed with small portions of MezSO (20 mL total). The filtrate was extracted with cyclohexane (3 X 50 mL). To the MezSO solution were added pyridine (0.7 mL) and carbethoxymethylene triphenylphosphorane (Alfa Products, Denvers, MA) (8 g, 22.4 mmol). Analysis of the solution after 1 h showed a single component by TLC (system EE), Rf 0.40 (Rf 0.06 for the 5’ aldehyde), and t R 10.1 min by HPLC in system H. The solution was partitioned between EtOAc (200 mL) and water (100 mL). The aqueous phase was extracted with EtOAc (2 X 100 mL). The EtOAc extracts were combined, washed with HzO (100 mL), dried (Na2S04),and evaporated in vacuo. The (25) Moroder, L.; Hallett, A.; Wunsch, E.; Keller, 0.; Wersin, G. Hoppe-Seyler’s 2.Physiol. Chem. 1976, 357, 1651.
Kappler et al. residue was dissolved in CH2Cl,-Me,C0 (4:1,100 mL) and filtered through a column of silica gel (5 X 10 cm) to remove excess of Wittig reagent. The column was washed with additional solvent to remove 17a. Volatiles were removed in vacuo to give a pale yellow foam, which was purified by preparative HPLC (CH,Cl,-acetone, 8812, one SiO, cartridge) to give 3.5 g (73.1%) of 17a as a white foam: ‘H NMR (300 MHz, CDCI,) 6 1.20 (t, 3 H, J = 7.2 Hz, CH,CH,), 1.39 and 1.62 (s, 3 each, CMe2),4.10 (q, 2 H, J = 7.2 Hz, CH,CH,), 4.82 (m, 1 H, H-4’), 5.13 (dd, 1 H, J3,,4i = 3.7 Hz, 53’2, = 6.2 Hz, H-3’), 5.54 (dd, 1 H, 52‘3’= 6.2 Hz, J1y = 1.9 Hz, H-2’), 5.82 (dd, 1 H, Ju,bt = 15.7 Hz, Jq,4, = 1.6 Hz, H-6’), 6.18 (d, 1 H, J182, = 1.9 Hz, H-l’), 6.94 (dd, 1 H, J5!,6, = 15.7 Hz, J5J,4j = 5.5 Hz, H-5’), 8.02 (m, 5 H, aromatic). Anal. (C24H25N506) C, H, N. N6-Benzoyl-5’-deoxy-5’(RS)-~-homocystein-S-yl-5’-( carbethoxymethyl)-2’,3’-0-isopropylideneadenosine (18a). To an aqueous solution (10 mL) of freshly prepared disodium Lhomocysteine (4 mmo1),22adjusted to pH 9 with 50% HI, was added 17a (0.96 g, 2 mmol) in THF (10 mL). HPLC showed the reaction to be complete in 0.5 h. The solution was evaporated. The residue was suspended in EtOH (20 mL) and centrifuged. The precipitate was washed with EtOH (2 X 10 mL). The EtOH solutions were combined and evaporated. The residue was dissolved in MeOH-H20 (4:6, 30 mL) and applied to a column of C-18 silica gel (2.5 X 15 cm). The column was washed with MeOH-H,O (4:6, 100 mL) to remove traces of amino acid, and 18a was then eluted with MeOH-H,O (4:1,100mL). Evaporation of volatiles in vacuo gave 18a (1.05 g, 82%) as a white powder homogeneous by HPLC in system H (tR 8.4 and 8.7 min; 5:4 mixture of diastereomers) and TLC (system D) (RF 0.50): ‘H NMR (300 MHz, CDC1,) 6 1.01 and 1.15 (t, 3 each, J = 7.2 Hz CH,CH3), 1.31, 1.34, 1.52, 1.54 (s, 3 each, CMe,), 3.82 and 4.04 (9, 2 each, CH,CH,), 6.15 (d, 1 H, J = 3.8 Hz, H-l’), 6.22 (s, 1 H, H-l’), 8.35, 8.48, 8.65, 8.80 (s, 1 each, H-2, H-8). Anal. ( C , S H ~ ~ N ~ O ~ S * ~C, . SH, H ~N, O )S. 5’-Deoxy-5’-(RS )-L-homocystein-S-yl-5’-(carbethoxymethyl)-2’,3’-0 -isopropylideneadenosine (18b). Disodium L-homocysteinate (3 mmol) was added to 17b (1.5 mmol)’s as described above for the preparation of 18a. After 0.5 h, THF was removed in vacuo, and the aqueous solution was adjusted to pH 5. The resulting precipitate was removed by centrifugation, and the clear supernate was applied to a column of C-18 silica gel (2.5 X 15 cm). The column was washed well with water to remove the amino acid, and then 18b was eluted with methanol-water (l:l,100 mL). Removal of volatiles in vacuo gave 18b as a white foam (0.61 g, 79.7%) homogeneous by TLC in (D), R, 0.12, and (R2), R f 0.48, and HPLC, tR 6.0 and 6.6 rnin (54.5 mixture of diastereomers in system H): ‘H NMR (300 MHz, CD30D) 6 1.11 and 1.22 (t, 3 each, J = 7.1 Hz, CH2CH3),1.38, 1.39, 1.58, 1.60 (s, 3 each, CMe2),6.13 (d, 1 H, J = 3.7 Hz, H-l’), 6.25 (d, 1 H, J = 2.0 Hz, H-l’), 8.21, 8.26, 8.27, and 8.32 (s, 1 each, H-2 and H-8). Anal. (C21H30N607S~H20) C, H, N, S. 5’-Deoxy-5’(RS )-I,- homocystein-S-yl-5’-(2-hydroxyethyl)-2’,3’-0-isopropylideneadenosine (19). Compound 18b (382 mg, 0.75 mmol) was dissolved in EtOH (50 mL), and a solution of Na13H4 in EtOH (1M, 3 mL) was added. The solution was kept at 60 OC for 2 h and then evaporated in vacuo. A solution of the residue in water was adjusted to pH 5 and applied to a column of C-18 silica gel (2.5 X 15 cm). The column was washed with water (100 mL) to remove amino acid. Water-methanol (6535; 100 mL) eluted 19, which was obtained as a white foam (132 mg, 38%) after evaporation of volatiles in vacuo. It was homogeneous on TLC (D), R 0.06, and on RPTLC (Rl),R, 0.33. Anal. (C1SH,8N606S.H20) H, N, s. The N-tert -Butyloxycarbonyl Derivative of 19, Methyl Ester 20. Compound 19 (117 mg, 0.25 mmol) was dissolved in DMF and triethylamine (40 pL, 288 pmol) and di-tert-butyl pyrocarbonate (65 mg, 298 pmol) were added. After 0.5 h at 22 OC, TLC analysis showed the reaction to be complete, Rf (D) 0.33 (ninhydrin negative) and RPTLC (Rl) R, 0.14. Volatiles were evaporated in vacuo. The residue was dissolved in MeOH (5 mL), and an excess of fresh ethereal CH,N2 was added. After removal of solvent in vacuo, the residue was chromatographed on a silica gel column (2.5 X 12 cm) in CHC13-MeOH (9752.5). Fractions containing the diastereomers of 20 were evaporated to a white foam (116 mg, 79.7%),which was homogeneous on TLC (B) R,
d,
Two Adducts of L-Methionine and P,y-Imido-ATP 0.14 and 0.17 (diastereomers) and (C) R, 0.44 and 0.48 and HPLC (H) tR 9.2 and 9.5 min (2:3 ratio of diastereomers). Anal. (C26H38Ns08S.CH,0H)C, H, N; S: calcd, 5.21; found, 4.73. 5'-Deoxy-5'-~-homocystein-S -yl-5'-[2-[ (p,y-imido)triphosphoroxy]ethyl]adenosine (Mixture of 5' Epimers, 16). A solution of 20 (100 mg, 0.17 mmol) and benzyl phosphate (130 mg, 0.69 mmol) in pyridine (5 mL) was evaporated. This was repeated twice. The residue was dissolved in pyridine and treated with dicyclohexylcarbodiimide (345 mg, 1.72 mmol). The mixture was kept at 22 "C for 48 h. Water (5 mL) was added, and after 1 h the solution was filtered and evaporated. A solution of the residue in AcOH (10 mL) containing P d black (200 mg) was shaken with Hz a t 40 psi for 18 h, filtered, and evaporated. To a solution of the residue in H,O-pyridine (l:l, 5 mL) was added NEt, (290 pL, 2.1 mmol). The solution was kept a t 22 "C for 10 min and evaporated. A solution of the residue in MeOH-H20 (1:9, 20 mL) was applied to a column of (2-18silica gel (2.5 X 15 cm, equilibrated with MeOH-HzO, 1:9). The column was washed with MeOH-HzO (1:9,100 mL) to remove inorganic phosphate. MeOH-H20 (l:l, 100 mL) eluted 21 (1600 ODzsounits, 62.7%), which was homogeneous on HVE (pH 7 3 , EMAMP 0.6. After removal of volatiles in vacuo, 21 (1550 ODzsounits, 0.M mmol) was dissolved in DMF (5 mL), and the solution was evaporated in vacuo. This was repeated twice. The resulting gum was dissolved in DMF (5 mL), and 1,l'-carbonyldiimidazole (80 mg, 0.5 mmol) was added. After 1 h a t 22 "C, MeOH (34 fiL) was added. The DMF was evaporated in vacuo after 0.5 h, and the residue was extracted several times with dry ether to remove most of the imidazole. The residual oil was dissolved in DMF (2 mL) and added to a solution of the tri-n-butylammonium salt of imidodiphosphate (0.5 mmol) in DMF (2 mL). The mixture was kept a t 22 "C for 18 h, when paper electrophoresis showed the reaction to be complete. A solution of imidazole in DMF (1 mL, 1 M) was added. The precipitate was removed by centrifugation and washed with a small volume of DMF. The DMF solutions were combined and evaporated in vacuo. The residue was dissolved in H 2 0 and applied to a column of c-18 silica gel (1 X 15 cm). This was washed with water (100 mL) to remove imidodiphosphate. Water-MeOH (3:1,100 mL) eluted 22 (880 ODzm units, 56.8%); removal of volatiles in vacuo gave 22 as a gum homogeneous by HVE a t pH 3.5 and pH 7.5. The gum was dissolved in 90% trifluoroacetic acid (TFA) (2 mL), and after 10 min a t 22 "C, the volatiles were immediately removed in vacuo. Toluene was evaporated twice from the residue to remove residual TFA. The residue was dissolved in buffer (1 mL) derived from 0.038 M NaHC0, and 0.025 M NaOH and adjusted to pH 10.4 with dilute NaOH. After 6 h at 22 "C, paper electrophoresis (pH 3.5) and PEI cellulose TLC (1 M LiCl) showed that hydrolysis was complete. The solution was chromatographed on a DEAE (HC03-) cellulose column (2 X 15 cm) with a linear gradient of Et3NH.HC03 (0-0.4 M, 2 L). Fractions containing 16 (580 OD,, units, 37.4%) were combined and evaporated. The residue was homogeneous on paper chromatography in systems (F) and (G), R, 0.33 and 0.40, respectively. The residue was dissolved in MeOH (500 pL), and 1 M NaI in MeOH (200 fiL, 5 equiv) was added. Acetone (20 mL) was added. The precipitate was washed with acetone and dried in vacuo over PzO5 to give the tetrasodium salt (23 mg) of 16 as a white powder: HPLC (system K), tR 10.8 and 12.3 rnin (2:3 ratio of 5' epimers); UV max pH 2,259 nm (e 15loo), pH 11,260 nm (e 15200); 'P NMR (121.46 MHz, DzO) b -11.79 ,
,
,
Enzyme'Studies. Preparations of M-2 and M-T were obtained as described previously.' Enzyme assays were conducted for 10 rnin a t 37 "C in a final volume of 0.1 mL containing 150 mM KC1-15 mM MgC12-5 mM dithiothreitol-50 mM TriseHCl, pH
Journal of Medicinal Chemistry, 1988, Vol. 31, No. 2 389 8.2.25 Each mixture was made up in duplicate. ~-[methyl-'~C]methionine (New England Nuclear Co., 54 Ci/mol) and MgATP were included at the levels specified below and in Table I, footnote a. A working enzyme solution was prepared freshly each day by 10-fold dilution of a stock solution with the buffer mixture against which it had been dialyzed and concentrated.8 Reactions were started by addition of 10 pL of working enzyme solution (9.5-10.5 X lo4 units of activity; 1 unit gives a V,,, with 2 mM ATP of 1pmol of product per min) and terminated by addition of 10 WL of 4 N HC104-lo mM L-methionine after immersing the solution in an ice bath. Each suspension was centrifuged and 50 pL of supernatant was applied to a 2.3-cm disc of phosphocellulose paper. Discs were washed as described,26then immersed in a toluene solution of phosphors and counted in a Packard liquid scintillation spectrometer (Model 2425). Controls were provided by incubations carried out in the absence of ATP. Reaction velocities were linear for a t least 30 rnin and were proportional to the amount of enzyme added a t the levels of enzyme activity employed. Inhibition studies were made with six to eight levels of MgATP or L-methionine in the range (0.5-4.0) X KM for each of two inhibitor levels that were in the range (1-10) X Ki and for control mixtures lacking inhibitor. Inhibitors were dissolved in the above pH 8.2 buffer solution prior to testing. Inhibition constants (Ki values) were obtained to within &15% from replots of inhibitor concentrations vs slopes or intercepts on the vertical (1/V) axis of double-reciprocal plots of velocity vs substrate level. All of the latter plots were linear, as were the replots. Typical double-reciprocal plots have been illustrated in a previous publication.'
Acknowledgment, This work was supported by Public Health Service Research Grant CA-11196 from the National Cancer Institute and by grants to the Institute for Cancer Research (USPHS Grants CA-06927 and RR-05539 and an appropriation from the Commonwealth of Pennsylvania). We thank Dr. Sidney Weinhouse, Temple University, for kindly providing the rat Novikoff ascitic hepatoma cells from which the M-T preparations were obtained. Registry No. 1, 84757-71-1; 2 (R-epimer), 111268-99-6; 2 (S-epimer), 111269-14-8; 3, 111269-00-2; 4, 111269-01-3; 5 (Repimer), 111269-02-4; 5 (S-epimer), 111269-15-9; 8 (R-epimer), 111323-50-3;8 (S-epimer),111269-16-0;9 (R-epimer), 111269-03-5; 9 (S-epimer), 111269-17-1; 10 (R-epimer), 111269-04-6; 10 (Sepimer), 111269-18-2;11 (R-epimer), 111269-05-7;11 (S-epimer), 111269-19-3; 12 (R-epimer), 111269-06-8; 12 (S-epimer), 11126920-6; 14 (R-epimer), 111291-44-2; 14 (S-epimer), 111269-30-8; 14.4Na (R-epimer), 111269-07-9;14.4Na (S-epimer), 111269-21-7; 14.EhN (R-epimer), 111291-45-3;14.Et3N (S-epimer),111269-31-9; 16 (R-epimer), 111269-32-0; 16 (S-epimer), 111269-34-2; 16.4Na (R-epimer), 111269-08-0; 16.4Na (S-epimer), 111269-22-8;16.Et3N (R-epimer), 111269-33-1; 16.Et3N (S-epimer), 111269-35-3; 17a, 111322-02-2;17b, 102044-62-2;18a (R-epimer), 111269-09-1; 18a (S-epimer), 111269-23-9; 18b (R-epimer), 111269-28-4; 18b (Sepimer), 111269-29-5; 19 (R-epimer), 111269-10-4;19 (S-epimer), 111269-24-0;20 (R-epimer), 111269-11-5;20 (S-epimer), 11126925-1; 21 (R-epimer), 111269-12-6;21 (S-epimer), 111269-26-2;22 (R-epimer), 111269-13-7; 22 (S-epimer), 111269-27-3; H-HcyOH*2Na,50615-55-9; C12CHCOOH,79-43-6; P-benzoyl-2',3'-0isopropylideneadenosine, 39947-04-1; methionine adenosyltransferase, 9012-52-6. (26) Liau, M. C.; Lin, G. W.; Hurlbert, R. B. Cancer Res. 1977,37, 427. (27) Hoffman, J. L. Methods Nzymol. 1983, 94, 223.
